102 M C K I N S EY Q UA RT E R LY Quantum’s Promise - - - - - - - - - - - - - - - Four Problems Quantum Computing Can Solve FOR BUSINESS LEADERS LOOKING to grasp the fundamentals of quantum computing (QC), we have identified four evolving QC capabilities that will emerge over the next three to ten years. All will be critical for solving com plex computational problems. Prime factorization could dramatically accelerate the factor ing of large integers compared with classical algo rithms. This process uses a sequence of steps known as Shor’s algorithm, which combines quantum interference with classi cal math routines to efficiently factor large integers and could break the math underpinning existing public-key cryptography. Although demonstrated for small numbers on quantum devices containing only a few quantum bits (qubits), prime factorization at scale will require fault-tolerant quantum computing, as well as advances in error correction and qubit scaling. That may take at least ten years. From the mid 2030s onward, prime factorization could become commonplace, necessitating urgent attention from companies today to ensure they secure their systems for a postquantum cryptography world. Quantum AI promises to harness quantum processors to either speed up existing AI training or enable fun damentally new learning paradigms, such as quantum neu ral networks that could analyze data far more efficiently than today’s AI systems. Full-scale deployments are still at least a decade away, but quantum- machine-learning (QML) algo rithms have been demonstrated in small-scale experiments. For example, researchers have devel oped a new liquid biopsy technique using QML that distinguishes between exosomes (microscopic particles released by cells) from cancer patients and those from healthy individuals by analyzing their electrical “fingerprints.” This approach could offer a faster, less invasive, and more cost-effective way to detect cancer. Quantum AI could also offer a synergistic feedback loop, as classical AI methods could improve quantum algo rithms and hardware control, potentially accelerating the maturation of quantum systems. Quantum optimization addresses complex combinatorial and numerical problems by leveraging quantum algorithms that explore vast solution spaces more efficiently than classical meth ods. While concrete field-specific examples are still emerging, early use cases in logistics and finance are demonstrating promising quantum optimization. For exam ple, financial-services companies could use QC for portfolio optimi zation, helping identify optimal asset allocations faster and more efficiently than classical computa tional techniques. This can allow a financial institution to determine the ideal investment mix within a portfolio and quickly alter its approach to respond to risks. In one example, Citi Innovation Labs has partnered with QC software company Classiq to explore oppor tunities for portfolio optimization. Exploratory quantum hard ware has also begun showing promise in accelerating heuristic algorithms to solve complex prob lems such as energy grid design or traffic flow management. - - - - - - - - - - - Broader commercial appli cations of heuristic optimization could emerge within five years, with large-scale deployments in the early 2030s. These applications will first leverage hybrid quantum–classical work flows before fully fault-tolerant machines become widely available. Quantum simulation can model complex systems, such as mol ecules and advanced materials, that are extremely chal lenging or impossible to simulate accurately on classical hardware because of the immense com putational resources required. For example, simulating the quantum behavior of just 50 qubits requires tracking more than one quadrillion states simul taneously. Quantum computers, by directly exploiting quantum principles such as superposition and entanglement, naturally rep resent these states, enabling the efficient and accurate simulation of quantum phenomena beyond the reach of classical computers. Quantum-simulation applica tions for chemistry, materials science, and drug discovery are already being deployed in small numbers and should reach scale within three years. Large-scale deployments are expected by the late 2020s to the early 2030s. This is contingent on develop ing error-corrected systems containing tens of thousands to hundreds of thousands of qubits.